Autonomous vehicles: Automotive and transportation disruption

It is extraordinary for a technology to have the opportunities and disruption potential that autonomous vehicles (AVs) will have over the next decades. Market opportunities worth trillions of dollars are forecast to emerge across transportation segments and different AV use cases. Tens of billions of dollars have been invested by high-tech companies, startups, auto manufacturers, their supply chains and many other companies—with more to come.

To give perspective, the table below summarizes the potential impact of AV technology. It is imperative to understand the difference between SAE Level 4 (L4) and Level 5 (L5) autonomous operation. L4 AVs can operate in specific areas that are usually defined by geographic boundaries. L4 AVs also include preprogrammed, fixed routes such as local bus routes, campus routes and similar trips—usually at low speeds. L5 AVs are autonomous vehicles that can operate anywhere within a country.

When AV technology is mature enough to deploy, the key advantages will be lower crash rates and fewer lives lost. AV software will avoid human drowsiness and driver distractions— especially with growing smartphone use. AV software senses nearby objects with faster reaction than humans. These factors will eventually make autonomous software a top choice.

Driverless vehicles can provide mobility at a lower cost, which will vastly expand the customer base for MaaS. Today, approximately one billion people have driver’s licenses. Over three billion people without a driver’s license could use AVs at affordable prices. Driverless MaaS will cost less than ride hailing, taxis and car ownership, depending on how many miles a person travels each year. The number of cars per household in developed countries is expected to decline as MaaS and AVs are deployed in volume. 

While the benefits are clear, companies operating in this space are currently racing to solve complex issues of software development, regulation issues, public opinion and cost of development.

AVs depend on successful development of multiple technologies including software, sensors and high-performance computing chips. AV software is the most complex technical component due to the vast number of potential driving events, although many are rarely seen. Rapid advancement of AI-based software technology is the key to create a safe response to unknown situations. Extensive testing on thousands of known driving situations is taking place, including rare events. In one example, Waymo is testing via virtual driving software that simulates AVs to find weaknesses and improve software capabilities. They drive 25,000 virtual cars simultaneously via computer simulation which covers billions of miles annually.

Multiple sensors are required to cover variations in lighting, weather and other issues, with camera, radar and lidar sensors being the most common. Camera and radar sensors  are already affordable and used for advanced driver-assistance.  Lidar is currently too expensive for AV deployment with a price range of $4,000–$50,000 for mechanical scanning lidars. Solid state chip-based lidar sensors are emerging and will eventually drop in price to a few hundred dollars. Chip technologies are also the solution for adding vastly higher computing power.

Multiple factors can delay AV deployment. Slow regulatory development is a potential barrier. Regulatory issues already have large variations; most AV testing is taking place in the few US states that have friendly regulations. California has given AV testing permissions to over 65 companies, with nearly half based in Europe and Asia. Over half of those are companies outside the auto industry.

There are many unknown AV variables that could create future problems. For example, low public trust and fatal AV crashes would delay AV deployment. Some crashes may be caused by companies releasing AV software before it is “smart” enough to be used on public roads. This would be a real problem as the capability of current testing in California varies by over 10 times between best and worst. The emerging use of simpler L4 use cases, described below, is expected to help build public trust in AVs.

Cybersecurity attacks that impact AV operation or result in a crash can delay AV deployment. Cybersecurity is among the toughest problems in any industry, including the auto industry. Cybersecurity is mandatory for any AV, and adding solutions to all connected cars requires significant investment.
Establishing safety rules, techniques and standards for developing and building AVs will be imperative as well. Many will be released in 2020.

Multiple factors could expedite AV deployment. Software technology breakthroughs are likely to have an impact—especially from an enormous AI software investment. The most likely breakthroughs are neural network technologies, since they are immature despite extensive investments in the last decade. Other AV accelerating factors could be chip or sensor technologies, or unexpected innovations.

An intriguing dynamic is the so-called network effect that is exhibited by any service where the benefits are proportional to the size of the user base. Examples include the Internet, phone social networks. Software platforms have large network effects, e.g., Microsoft Windows, Apple iOS and Android. Network effect for AV software is expected in a different way; the goal being to handle any or as many driving situations as possible. The more AVs that are deployed and the more miles that are driven,  the more driving situations are learned. This new knowledge is quickly updated, and new software capabilities are downloaded to every fleet vehicle. The result is that large AV fleets will quickly gain an advantage in AV software capabilities versus later-starting and smaller AV fleets. Waymo and others may be on this track and will reach acceptable AV safety before their competitors.

AV deployment will proceed incrementally based on when AV software can meet overall safety requirements for AV use cases. This means that AV use cases with low traffic complexity and low speed will deploy first. The following table summarizes the main AV use cases.

Fixed route passenger transportation is among the simplest AV use cases. Fixed routes with low to moderate speed requirements have completed hundreds of tests and some deployments. Closed-venue applications such as airports, universities and similar settings are deploying as well.

AVs for last-mile goods delivery is another early AV use-case. This includes meal and grocery deliveries and increasing e-commerce deliveries. Current deliveries are via human-driven vehicles including bikes. Last-mile goods AVs have cost and other advantages. Last-mile goods AVs range from small, inexpensive sidewalk AVs to an intermediate size, and also to a car-based version. Sidewalk AVs have experienced the most progress with significant restaurant and grocery delivery tests and deployment in closed venues. 

Autonomous trucks are used for hub-to-hub trucking, which are simple AV use cases. Routes for hub-to-hub trucking are mostly highways, which current AV software technology can manage. First and last miles of hub-to-hub trucking are done by truck drivers or via teleoperation by a remote driver. The truck driver shortage in most areas—especially long-haul trucks—makes this a desirable AV segment.

L4 driverless ride hailing or robo taxis have received the most attention and investments among AV segments. It is more complex than fixed route AVs, last-mile goods AVs and hub-to-hub autonomous trucking. Robo taxis operate in environments with more traffic complexity, including areas with many pedestrians, which adds difficulties to AV software development and testing. These AV complications mean that robo- taxis will deploy on a city-by-city basis, initially covering part of a city and expanding with experience. 

Personal L4 AVs are autonomous vehicles for individual use within specified areas. Personal L4 AVs are much harder to develop and test. Personal AV deployment is likely to follow the robo-taxi city-by-city pattern but is limited to specific metro areas. For use in another region, the software would require wireless updates. Since most people drive over 95% in their home town, this would be valuable to a large portion of vehicle owners. Personal AV deployment will lag robo-taxis by over five years.

Personal L5 AVs can drive anywhere within a country. This is the ultimate in AV software development and testing complexity. L5 AV availability is uncertain, and there is doubt whether L5 AVs will ever happen. New technologies and innovations are required with substantial AI advancements as a key technology for future success. 

AV technology is emerging and will have major disruptive impact on automotive and many other industries. The disruption will be amplified by concurrent introduction of battery electric vehicles and expansion of MaaS business models. These three win-win-win technologies will radically change automotive, transportation and other industries. The timeline is fuzzy, but it is no longer a question of if it will arrive, but rather when and how fast.

About IHS Markit (www.ihsmarkit.com)
IHS Markit (NYSE: INFO) is a world leader in critical information, analytics and solutions for the major industries and markets that drive economies worldwide. The company delivers next-generation information, analytics and solutions to customers in business, finance and government, improving their operational efficiency and providing deep insights that lead to well-informed, confident decisions. IHS Markit has more than 50,000 business and government customers, including 80 percent of the Fortune Global 500 and the world’s leading financial institutions. Headquartered in London, IHS Markit is committed to sustainable, profitable growth.
 
Automotive offerings and expertise at IHS Markit span every major market and the entire automotive value chain—from product planning to marketing, sales and the aftermarket.  For additional information, please visit www.ihsmarkit.com/automotive or email [email protected]
  
IHS Markit is a registered trademark of IHS Markit Ltd. and/or its affiliates. All other company and product names may be trademarks of their respective owners © 2019 IHS Markit Ltd. All rights reserved.